A viral video arrives during breaking news with a simple promise: it shows what happened. Your job is to decide whether that promise holds. The clip may be real, old, edited, mislabeled, or from another country. Verification is not a single trick; it is a short routine repeated under pressure. The checklist below is designed for beginners and for experienced reporters who need to slow their hands for five minutes.

Begin with the claim, not the footage

Before you inspect a single frame, write down what the video is supposed to prove. “Flooding in the city center after the storm” is a claim. “A clip posted at 18:42 by an account called @localweather” is a fact about the post. Keep those categories separate. Screenshot the post, save the URL, note the account name, the caption, the upload time, and any visible replies or community notes. If the post disappears later, your notes become the record.

Ask a simple question: what exactly would this video need to show for the claim to be true? A street sign, a known building, a particular bridge, a uniform, a vehicle plate, a weather condition. That answer shapes the rest of your verification. It also stops you from being impressed by motion, smoke, and shouting when the clip may be from a different event entirely.

Trace the source backwards

Most viral videos are reposts. Your task is to find the earliest available instance, not the most popular one. Search the caption’s distinctive phrases, the account name, and likely local-language keywords. Use keyframes from the video for reverse image searches; pick frames with buildings, signs, vehicles, or unusual terrain. If the clip is short, extract several frames because one frame may match an unrelated stock image or news photo.

Then inspect the account that posted it. When was the account created? What did it post before this event? Does it claim to be in the place it is describing? Does it mix original footage with recycled clips? A local resident posting from a known street carries different weight from an anonymous aggregator that posts disasters from three continents. Neither is automatically trustworthy, but the context tells you how much work remains.

If you can contact the uploader, do so. Ask where and when they filmed, whether they have the original file, and whether they saw the event themselves. Keep the questions plain. “Can you send the original clip with its file details?” and “What is the nearest landmark?” are more useful than asking them to confirm your theory.

Read time carefully: timestamps, metadata, and context

A timestamp on a post is not proof of when the video was filmed. Platforms display upload times in your time zone, the account may have scheduled the post, and the clip itself may be years old. File metadata can help, but it is often stripped when a video is downloaded, edited, or forwarded through messaging apps. Treat metadata as a clue, not a verdict.

Look for time markers inside the footage: shadows, clocks, shop shutters, streetlights, weather, traffic patterns, and live broadcasts. If the claim says “now” during a storm, compare the clip with current weather reports and local live streams. If the clip shows bare trees while the claim describes summer, that is a problem. If audio and video seem slightly out of sync, or if the clip cuts just before a useful detail, slow down and examine the joins.

What metadata can and cannot tell you

Creation dates and device details can support a timeline, but they can also be wrong, edited, or missing. A file with no metadata is not automatically fake. A file with neat metadata is not automatically genuine. Use metadata alongside visual evidence, source history, and independent confirmation.

Geolocate the scene

Geolocation means placing the footage in the real world. Start with the obvious: street signs, shop names, road markings, bus stops, bridges, hills, cell towers, and the shape of buildings. Compare those details with maps, satellite imagery, and street-level views. Check the direction of sunlight against the claimed time of day. A shadow falling the wrong way can expose a false location or a false time.

Work from wide to narrow. Identify the city or region first, then the neighborhood, then the exact corner. Use several frames because a single frame may hide the detail you need. Remember that footage can be mirrored, cropped, or stretched. If a sign reads backwards, flip the image and search again. If the claim places the video in one country but the plates, language, and emergency service uniforms point elsewhere, note the discrepancy and keep checking. Clues are not proof until they fit together.

Cross-check with independent evidence

Two accounts sharing the same clip are not two sources. Look for separate material: a different angle, a longer recording, a local journalist on the ground, an official statement that matches specific details, or a live stream showing the same location at the same time. The strongest confirmation usually comes from details that were not easy to fake: the same damaged roof, the same row of parked cars, the same plume of smoke seen from another street.

Ask what would prove the clip wrong. If the event is described as a current explosion, does the local fire service report an incident at that time? If the video claims a famous landmark, does the landmark look correct from that angle? If no independent evidence exists, say so plainly. “A video circulating online appears to show…” is honest. “This video shows…” is a promise you may not be able to keep.

Publish with caveats and a chain of custody

Before publication, keep a simple verification note: where the clip was found, when it was posted, what the uploader said, which frames you used for geolocation, and which details remain unconfirmed. Preserve screenshots and links with timestamps. If you use the video, avoid cropping out context that changes its meaning. Do not use a single clip as sole proof of casualty figures, cause, or intent. Those claims need separate reporting.

Write captions that match the evidence. If the location is confirmed but the date is not, say that. If the date is confirmed but the source is anonymous, say that. Corrections are part of the job; a quick update is better than a stubborn error.

A five-minute checklist for the desk

1. Write the claim in one sentence and separate it from the post’s caption.
2. Save the URL, account details, upload time, and screenshots.
3. Find the earliest available version of the clip.
4. Check the uploader’s history and contact the source if possible.
5. Extract keyframes and run reverse image searches.
6. Compare post time, claimed time, shadows, weather, and local live streams.
7. Geolocate using signs, buildings, roads, and satellite or street-level maps.
8. Look for independent footage, local reporting, or official confirmation.
9. Label what is confirmed, what is unconfirmed, and what is unknown.
10. If you cannot complete the routine, publish the claim as unverified or hold it.

The goal is not to kill a compelling video. The goal is to know what you actually have before your audience does.

Photo: StockSnap / Pixabay